**** BEGIN LOGGING AT Fri Nov 06 02:59:57 2009 Nov 06 06:12:07 good morning Nov 06 07:23:58 hi, sorry to rehash yesterday's topic, but I want to raise the question of the location of mysqld.pid again Nov 06 07:24:14 yesterday we agreed it should be in /var/run/mysql (which I created in the recipe). Nov 06 07:24:48 however this is not a persistent situation as /var/run links to /var/volatile/run and this is re-created each time Nov 06 07:25:00 so i think the options are: Nov 06 07:25:48 - create the pid dir in the init.d script (or somewhere else) Nov 06 07:26:45 you could put it in /var/lib Nov 06 07:26:59 - put the dir somewhere else. I've seen locations like /tmp, /tmp/mysql and /usr/local/mysql Nov 06 07:27:14 van/lib seems also good to me Nov 06 07:27:38 not really an idea what is standard Nov 06 07:27:56 most distros use /var/lib in such cases Nov 06 07:28:03 ok, will do that Nov 06 07:28:39 good night Nov 06 07:28:50 * khem -> sleep () Nov 06 07:28:56 night Nov 06 07:41:24 03Frans Meulenbroeks 07org.openembedded.dev * r189a5b7f76 10openembedded.git/recipes/mysql/ (files/my.cnf files/mysqld.sh mysql_4.1.22.bb): Nov 06 07:41:24 mysql: moved pid dir from /var/run/mysql to /var/lib/mysql as /var/run is created upon boot Nov 06 07:41:24 (and so var/run disappears) Nov 06 07:41:25 added restart to init.d script Nov 06 07:55:57 03Sebastian Spaeth 07shr/merge * r71a8d0e5ee 10openembedded.git/recipes/tasks/ (task-shr-feed.bb task-shr.bb): Nov 06 07:55:57 task-shr: move mokomaze from feed into full image Nov 06 07:55:57 Signed-off-by: Sebastian Spaeth Nov 06 08:01:09 eFfeM: why do you want the .pid file to persist across reboots? Nov 06 08:01:27 03Frans Meulenbroeks 07org.openembedded.dev * r35cc93da88 10openembedded.git/ (11 files in 3 dirs): mysql5: created recipe Nov 06 08:02:06 pb__, i don't want the pid to persist, the problem is that the pid file is written by user mysql who has no write permission in /var/run Nov 06 08:02:24 yesterday it was suggested I would create a dir /var/run/mysql during install and write in there Nov 06 08:02:32 but that dir is gone after the reboot Nov 06 08:02:51 that was the point i wanted to raise above Nov 06 08:02:54 ah, I see Nov 06 08:03:11 right, fair enough I guess Nov 06 08:03:31 actually maybe it would have been better to create the mysql dir in /etc/init.d/mysql as that one still runs as root iirc Nov 06 08:08:27 hi all, and sorry I screwed git up: Nov 06 08:08:40 I locally renamed org.openembedded.org to dev and "git push'd" now openembedded.org has a "dev" branch. Nov 06 08:08:43 yeah. that does sound a bit better Nov 06 08:09:17 Can we remove that branch again, please? Nov 06 08:09:40 spaetz: one of the admins will have to do that on the server. Nov 06 08:09:55 pb___: ok, so send a mail to cbrake or so? Nov 06 08:10:10 yeah, that sounds like your best option Nov 06 08:10:15 k, thanks Nov 06 08:10:30 afaik, cbrake is the main man for git, though there is a list of all the melo administrators on the wiki Nov 06 08:11:09 http://wiki.openembedded.org/index.php/Infrastructure Nov 06 08:20:37 In my recipe I trying to create a link from /usr/bin/netconf -> /path/to/my/script.sh. I do that with "ln -s ${D}/qmdc/scripts/netconf.sh ${D}${bindir}/netconf". The package builds ok and when I look at the content of the ipk everything looks right. But when I include the package in my image the link is not being created. My recipe and the output from dpkg-dep can be seem here: http://pastebin.com/m5f089a3c Nov 06 08:21:06 Isn't the linking done correctly? Nov 06 08:26:10 jovox: you've got too many ${D}s in there Nov 06 08:26:48 ln -s "/qmdc/scripts/netconf.sh" "${D}${bindir}/netconf" Nov 06 08:37:25 hi, does anyone experience this: http://bugs.openembedded.org/show_bug.cgi?id=5332 ? Nov 06 08:50:41 pb__ Thanks, I will try that Nov 06 09:12:46 jo Nov 06 09:14:00 morning woglinde Nov 06 09:15:30 hi ph5 Nov 06 09:18:02 morning all Nov 06 09:18:07 morning all Nov 06 09:18:52 hi all Nov 06 09:19:14 hi Nov 06 09:19:27 * eFfeM wants to join the morning roll call :-) Nov 06 09:19:37 heh Nov 06 09:19:45 hi eFfeM, rp, ph5, woglinde, th1, all Nov 06 09:20:25 groan, my poxy mailserver is out of disk space again. how does that keep happening? Nov 06 09:20:27 * pb___ stomps around Nov 06 09:22:24 oh well, breakfast time now Nov 06 09:23:07 hi rp and pb Nov 06 09:23:25 hi effem Nov 06 09:23:36 effem mythtv works now? Nov 06 09:24:37 woglinde: more or less, still ironing out the last issues (e.g. add mythweb) Nov 06 09:24:53 and need to do some more testing Nov 06 09:25:24 but got backend working on beagle (with my desktop as frontend) Nov 06 09:25:51 haven't tested frontend on OE yet, will do so on my beagle probably later today Nov 06 09:25:56 effem *g* Nov 06 09:26:16 at least making waves :-) Nov 06 09:26:57 hm let me see if I need a travel adapter in the uk Nov 06 09:27:14 eFfeM: and it's viable performance wise? Nice work. Nov 06 09:27:23 he djwillis ;) Nov 06 09:27:40 hm there was one think I hate on mythtv mysql Nov 06 09:27:44 hola Nov 06 09:27:52 woglinde: from whare, UK will need a 3pin travel adaptor but it's all 220/240v Nov 06 09:28:07 XorA: howdi Nov 06 09:28:23 morning xora Nov 06 09:28:53 woglinde: you dont even really need the adapter, a byro will persuade .eu adapters into UK sockets Nov 06 09:29:01 hm Nov 06 09:29:02 http://www.netzgeraet.de/index.html?reisesteckernetzteile_stromadapter_fuer_urlaub_und_reisen.htm Nov 06 09:29:26 byro? Nov 06 09:29:36 woglinde: a pen Nov 06 09:30:00 aeh? Nov 06 09:30:10 cost not that much Nov 06 09:30:42 eu->uk costs about 20p Nov 06 09:35:59 ao2: this won't help you, but I just did a rm -rf tmp, bitbake ncurses (armv5te on x86_64), no complaints. both oe and bitbake git head. Nov 06 09:41:14 DJWillis: can't really judge the performance, I have a card with a hw mpeg encoder, then the backend works fine Nov 06 09:41:34 you probably do not want to do transcoding Nov 06 09:42:50 effem hm? dsp on beagleboard should be fast enough Nov 06 09:43:15 guess so, but not sure if there are good libs for it yet, didn't really dig into that area Nov 06 09:43:50 pH5, I am also using OE head and bitbake 1.8.16. I can try removing tmp and rebuild, just to keep the CPU busy while waiting for comments on the bug report :|. Thanks for answering, btw. Nov 06 09:44:02 and actually I only have two usb receiver cards, one analog hauppauge wintv usb pvr2, and one dvb-t card which has no linux drivers (and indoor antenna for dvb-t is not good enough here anyway) Nov 06 09:44:43 hu which dvb-t card has no linux driver? Nov 06 09:44:55 main worry i have with beagle is usb performance Nov 06 09:45:02 ao2 use newer bitbake Nov 06 09:45:13 for .dev Nov 06 09:45:37 woglinde, will it work with python2.4 ? Nov 06 09:45:50 install newer pyton Nov 06 09:46:08 it means "no" :) Nov 06 09:46:15 dont know Nov 06 09:46:24 but why you have to stick with python 2.4? Nov 06 09:46:37 newest bitbake needs python 2.6 which isn't even in lenny-backports :( Nov 06 09:46:52 I had to compile it from source from experimental Nov 06 09:46:55 th1 install from ubuntu-repo Nov 06 09:46:59 no pain Nov 06 09:47:09 IC Nov 06 09:47:12 i'll try that next time :) Nov 06 09:47:40 deb http://ppa.launchpad.net/doko/ubuntu intrepid main Nov 06 09:47:44 I think was it Nov 06 09:47:54 after installing you can disable the entry Nov 06 09:48:20 it's more or less the same as I built from source Nov 06 09:48:26 nope Nov 06 09:48:35 how do you know ;) Nov 06 09:48:38 building from source cost more time Nov 06 09:48:46 yeah but now I already built them into debs Nov 06 09:48:50 so next time I can just install them again Nov 06 09:48:51 ;) Nov 06 09:49:02 XorA: do you have any Sharp CE-170TS Serial Cable ? Pavel Macheck is still lacking it. Nov 06 09:49:03 and these are built for lenny not ubuntu Nov 06 09:49:55 ant_work: I do, but that means I can not work on collie kernel anymore Nov 06 09:50:02 no no Nov 06 09:50:19 iirc Laibsch had some and I asked him to send one to Pavel Nov 06 09:50:25 long ago... Nov 06 09:51:11 if pavel can solder I have a spare connector somewhere Nov 06 09:51:35 well, if one sorts out it would be easy to give it to hrw or woglinde, near 'neighbours' Nov 06 09:52:03 hm xora wouldnt make it Nov 06 09:52:04 woglinde: no idea where I put the dvb-t card, too much junk around here (please don't tell my wife :-) she'll only say she already knew) Nov 06 09:52:13 effem *g* Nov 06 09:52:16 but it was one of those el cheapo ebay cards for eur 15 Nov 06 09:52:23 might have given it to one of the kids Nov 06 09:52:24 woglinde: ????? Nov 06 09:52:36 xora wasnt you ill? Nov 06 09:52:39 hm Nov 06 09:52:48 but looking for it i found an avermedia vartv usb2.0 plus, think it is mjpeg or so Nov 06 09:52:49 yeah not at OEDEM Nov 06 09:52:57 ah Nov 06 09:52:58 okay Nov 06 09:52:59 sorry Nov 06 09:53:00 and a technisat airstar dvb-t usb which i never tried Nov 06 09:53:19 woglinde: was someone expecting me? Nov 06 09:53:36 shouldnt I take something for thesing with me? Nov 06 09:54:08 guess the M in eFfeM stands for Messy Nov 06 09:55:22 I have a Zaurus Serial Cable too... (in germany) Nov 06 09:58:45 crappy computers... they never think... :) Nov 06 09:59:48 woglinde: Ill post it direct instead Nov 06 09:59:59 woglinde: I dont think lrg is going either Nov 06 10:00:33 xora ah okay Nov 06 10:01:08 * XorA needs to post some zauruses next week anyway Nov 06 10:05:46 woglinde: I'm not going, I didn't even feel well enough to go to the pub with XorA last night either :( Nov 06 10:06:05 XorA: you electr. engineer? pls. solder the cable :) Nov 06 10:06:49 lrg ;( Nov 06 10:06:52 ant_work: burglarise zecke :-D Nov 06 10:07:04 ant come to berlin Nov 06 10:07:10 I can grab it for you Nov 06 10:08:10 \o/ my mass angstrom feed update seems to have worked Nov 06 10:21:44 hm now downloading the osm map for england Nov 06 10:21:59 good morning. are the instructions for using a pre-built toolchain supposedly current. ASSUME_PROVIDED doesn't seem to be having any effect Nov 06 10:22:26 mike_cw somthing is in the wiki and the manual Nov 06 10:22:45 i'm following the manual at the moment - i'll check the wiki Nov 06 10:23:27 i wanted to use a toolchain built with meta-toolchain purely to save time on full builds Nov 06 10:23:48 lrg: wow, you must be feeling bad Nov 06 10:25:56 damn why openstreetmap changed the cool download feature Nov 06 10:26:32 ah okay Nov 06 10:37:26 pb__: not long ago we talked about fixing the -dev packaging for an opie plugin. You suggested to add the opie/plugin only once in an opie file...you meant in palmtop.bbclass isn't? Nov 06 10:37:53 wow! osm has come leaps and bounds, (pretty sure this is what i looked at years ago and seemed like a project with no hope of getting anywere) Nov 06 10:38:01 ant_work: yeah, probably Nov 06 10:38:11 I'm not all that au fait with the opie packaging Nov 06 10:38:24 Neither. But I see there is FILES_${PN}-dbg Nov 06 10:38:58 I'll try adding -dev Nov 06 10:38:59 mango? Nov 06 10:39:05 mango try osm with navit Nov 06 10:39:12 morning folks Nov 06 10:39:15 mickeyl: good morning! Nov 06 10:39:19 thats best routing/gps/stuff for free Nov 06 10:39:21 he mickeyl Nov 06 10:39:46 woglinde: abount navit..does sound/voice now work? Nov 06 10:40:16 pb__: hey. i think i finally figured out how to travel from LCY to Cambridge. It's DLR [Julbilee] -> Bank [Northern] -> King's Cross [National] -> Cambridge :) Nov 06 10:40:19 does that look sane? Nov 06 10:40:38 mickeyl: yah, that sounds about right. Nov 06 10:40:42 woglinde: was OE problem, binding espeak/pulseaudio..don't remember Nov 06 10:40:48 mickeyl: you could have saved yourself some time by looking on the wiki :-) Nov 06 10:40:48 excellent. one question though, how does it work with fares there? Nov 06 10:40:50 ant didnt need it yet Nov 06 10:40:54 aargh, really? Nov 06 10:40:57 heh Nov 06 10:41:03 well, finding out on my own is better for learning ;) Nov 06 10:41:11 http://wiki.openembedded.org/index.php/Oedem/2009#Travel Nov 06 10:41:14 heh, right Nov 06 10:41:18 at we have a project with the main navit dev Nov 06 10:41:20 *g* Nov 06 10:41:30 cp15? Nov 06 10:41:42 dont know his nick Nov 06 10:41:47 he crofton Nov 06 10:42:04 the DLR and underground have integrated fares. let me check what ticket you need from LCY. Nov 06 10:42:07 gm Nov 06 10:42:11 pb__: ah great, thanks Nov 06 10:42:29 morning Crofton Nov 06 10:43:23 hey Crofton safely in Englandshire? Nov 06 10:43:27 damn adhoc stuff Nov 06 10:43:29 http://www.tfl.gov.uk/assets/downloads/standard-tube-map.pdf is the london transport map Nov 06 10:43:53 yeah Nov 06 10:43:58 crashing after transfaring 2 Mb Nov 06 10:43:59 In "Milton Keynes" Nov 06 10:44:06 lcy is in zone 3, so you need a zone 1-3 ticket Nov 06 10:44:07 Crofton: ewwww hehe Nov 06 10:44:13 batteries died and no car charger Nov 06 10:44:17 this place is weird Nov 06 10:44:23 pb__: wont a return to Cambridge from LCY work? Nov 06 10:44:27 seems like a zombie town Nov 06 10:44:35 crofton *g* Nov 06 10:44:45 pb__: ok, thanks. Nov 06 10:44:51 pb from the youth hostel and to? Nov 06 10:45:05 XorA: does LCY have a mainline station? Nov 06 10:45:11 hm from rail station to youth hostel Nov 06 10:45:31 mickeyl: cash price appears to be £4, or it's £2.70 on an oyster card Nov 06 10:45:42 pb__: did they stop selling integrated tickets from undergrounds these days? Nov 06 10:46:04 pb__: last I was in london I used my oyster card Nov 06 10:46:05 XorA: I don't think they sell them on the DLR. Nov 06 10:46:13 pb__: curse the DLR Nov 06 10:46:20 pb__: worse train I ever been on :-D Nov 06 10:46:30 dlr ticketing is very basic, it's set up primarily for oyster Nov 06 10:46:39 oayster? Nov 06 10:46:48 mickeyl: then, when you get to kings cross, you need to buy a mainline ticket to cambridge. Nov 06 10:46:55 mickeyl: oyster is the smartcard pay-as-you-go thing Nov 06 10:47:07 ah, interesting Nov 06 10:47:14 modern technologies Nov 06 10:47:22 * mickeyl prefers paper tickets ;) Nov 06 10:47:25 heh Nov 06 10:47:34 yeah, paper ticket probably makes more sense if you are only making one return journey Nov 06 10:48:10 what will the mainline trip rougly cost? Nov 06 10:48:12 chances are if you went oyster you wouldnt be able to charge it anyway :-( Nov 06 10:48:19 stupid system spends half its life crashed Nov 06 10:48:20 heh Nov 06 10:48:39 mickeyl: I think it's about £20-£25 for a return. let me check. Nov 06 10:49:11 (just need a rough idea how much many to draw when i'm at LCY) Nov 06 10:49:16 err Nov 06 10:49:18 s/many/money/ Nov 06 10:49:21 can't write today Nov 06 10:49:44 mickeyl: you can pay with visa anyway, I wouldn't bother taking cash out for the train Nov 06 10:50:07 pb__: good point! Nov 06 10:50:24 ah..atm some images are broken by sqlite3 failing. Seems a missing dep from tcl-native. Isn't that dep optional? Nov 06 10:50:53 ant hm in the latest there are some headers generated Nov 06 10:50:54 i guess coldham park is not in walking distance from the train station? Nov 06 10:50:54 you used to need cash for the tube, but they have had ticket machines that accept cards for several years now. Nov 06 10:51:03 maybee its only a configure option Nov 06 10:51:07 mickeyl: depends if you like walking. Nov 06 10:51:10 it's about 45 mins on foot Nov 06 10:51:10 woglinde: Hi, pls can you take a look at those two? http://ynezz.ibawizard.net/oe/0002-geodegx-Fix-glibc-configure-error.patch http://ynezz.ibawizard.net/oe/0001-geodegx-Linux-2.6.31.4.patch Nov 06 10:51:11 hehe Nov 06 10:51:20 woglinde: I got (host openembedded.org[188.40.83.200] said: 451-82.208.49.253 is not yet authorized to deliver mail from Nov 06 10:51:30 pb__: ah well, with luggage that's a bit more than I appreciate :) Nov 06 10:51:36 mickeyl: see the wiki for walking directions and other travel options :-) Nov 06 10:51:41 awesome, thanks Nov 06 10:51:47 * mickeyl praises organization Nov 06 10:53:17 I would offer to pick you up from the station, but I need to babysit Emma between 17:00 and 19:00 this afternoon so I don't think that will work. Nov 06 10:54:37 yeah, that might be around the time i arrive (when everything works ;) If i get lost, I'm sure there's an app on my iPhone to get me back on track Nov 06 10:55:17 mickeyl, make sure the battery is charged :) Nov 06 10:56:18 Crofton: hehe, yeah. doing that now Nov 06 10:56:33 it lasts for barely more than one day Nov 06 10:56:43 so it needs daily charging anyways Nov 06 10:56:59 mickeyl: heh, yes, no doubt. feel free to call me if you do get lost, but I imagine you will be ok; cambridge is not all that big :-) Nov 06 10:57:09 pb__: right Nov 06 10:57:49 ynezz is geode really working with i586? Nov 06 10:57:58 ynezz is geode support gone from glibc? Nov 06 10:58:28 are see the configure msg now Nov 06 10:58:47 it looks like that Nov 06 10:58:53 and it's working on my box Nov 06 10:58:54 pb__: ah btw., i texted you from my new number, had to change providers. should arrive soonish Nov 06 10:59:00 ynezz cool Nov 06 10:59:07 could please sent it to the oe-ml Nov 06 10:59:12 I will check it in Nov 06 10:59:12 I tried Nov 06 10:59:15 hm Nov 06 10:59:23 problems? Nov 06 10:59:28 11:51:17 < ynezz> woglinde: I got (host openembedded.org[188.40.83.200] said: 451-82.208.49.253 is not yet authorized to deliver mail from Nov 06 10:59:32 maybee some played with the dns entry? Nov 06 10:59:51 cbrake mentioned something about maintance Nov 06 11:00:21 maybe it went now in Nov 06 11:00:21 "451 not yet authorized" sounds more like greylisting Nov 06 11:00:30 my qeue is empty Nov 06 11:00:51 mickeyl: thanks, got it Nov 06 11:01:11 ynezz send from other host or other mailrelay? Nov 06 11:01:27 ynezz hm wait Nov 06 11:01:31 I will do it Nov 06 11:01:38 cost me only 5 minutes I guess Nov 06 11:03:46 thanks Nov 06 11:06:38 ynezz done :) Nov 06 11:06:42 gogo cia Nov 06 11:06:44 03Petr Štetiar 07org.openembedded.dev * r7e28606c94 10openembedded.git/conf/machine/ (geodegx.conf geodelx.conf): (log message trimmed) Nov 06 11:06:44 geodegx: Fix glibc configure error Nov 06 11:06:44 Fix glibc configure error: Nov 06 11:06:44 | configure: running configure fragment for add-on nptl Nov 06 11:06:44 | checking sysdep dirs... configure: error: The geode subspecies of i486 is not supported. Nov 06 11:06:46 | FATAL: oe_runconf failed Nov 06 11:06:50 Signed-off-by: Petr Štetiar Nov 06 11:06:54 03Petr Štetiar 07org.openembedded.dev * r2c92869662 10openembedded.git/ (4 files in 4 dirs): Nov 06 11:06:55 geodegx: Linux 2.6.31.4 Nov 06 11:06:57 Signed-off-by: Petr Štetiar Nov 06 11:06:59 Signed-off-by: Henning Heinold Nov 06 11:08:00 that was quick :) Nov 06 11:10:12 damn navit Nov 06 11:10:14 you git Nov 06 11:10:20 aeh yo git make easy Nov 06 11:12:01 you said git and easy in the same sentence.. although I didn't make any *sense* of that sentence ;) Nov 06 11:12:31 th1 then still try svn Nov 06 11:12:41 git makes many thinks really easy Nov 06 11:13:08 mickeyl: mainline train ticket is £29 according to firstcapitalconnect.co.uk Nov 06 11:13:38 pb__: ok, thanks Nov 06 11:13:44 I trust First about as far as I can throw them so that figure should probably be treated with some caution, but at least it gives you a rough idea Nov 06 11:14:02 yeah, that's good enough for an estimate Nov 06 11:14:20 that was for a return, outbound today and returning sunday evening Nov 06 11:17:18 right. I'm leaving on monday though Nov 06 11:17:40 the flight would have been much more expensive otherwise Nov 06 11:18:26 oh right Nov 06 11:18:42 I don't imagine that will change the price much Nov 06 11:18:48 yea Nov 06 11:19:29 yeah, same price for monday Nov 06 11:20:46 mickeyl: good choince not getting a UK train on a sunday Nov 06 11:21:00 mickeyl: we probably lose you for months Nov 06 11:21:09 XorA: heh, that tough? Nov 06 11:21:27 mickeyl: its a disaster normally Nov 06 11:21:42 mickeyl: they just dont seem to be able to schedule around engineering works Nov 06 11:22:05 XorA: right, read a lot of warnings about that on their websites Nov 06 11:22:10 at least they're warning you Nov 06 11:22:28 mickeyl: they publish alternative timetables, then just ignore them normally Nov 06 11:22:36 03Frans Meulenbroeks 07org.openembedded.dev * rb441c5e7b9 10openembedded.git/recipes/mythtv/mythtv_0.21+0.22rc2.bb: mythtv: added mc.sql; introduced database package for that one (similar to debian) Nov 06 11:22:37 hehe Nov 06 11:22:47 03Frans Meulenbroeks 07org.openembedded.dev * rbc6d9ef31b 10openembedded.git/: Merge branch 'org.openembedded.dev' of git.openembedded.org:openembedded into org.openembedded.dev Nov 06 11:23:17 rwhitby: ping Nov 06 11:23:33 mickeyl: syn|ack Nov 06 11:23:55 rwhitby: heya. i can't seem to log in to melo after the reboot. could you check whether my key is still present? Nov 06 11:25:29 mickeyl: mickey@ds9 key is there Nov 06 11:25:36 right, that's it Nov 06 11:25:46 hmm Nov 06 11:26:02 aha: User mickey not allowed because account is locked Nov 06 11:26:08 d'oh Nov 06 11:26:23 any particular reason for that? Nov 06 11:26:36 dunno. how do I unlock an account on ubuntu 8.04 ? Nov 06 11:27:15 aha, you have a ! instead of an * Nov 06 11:27:31 google says sudo passwd -u username Nov 06 11:28:10 mickeyl: try now Nov 06 11:28:33 works again Nov 06 11:28:42 i should find out how locked it though Nov 06 11:28:48 err Nov 06 11:28:48 who Nov 06 11:29:47 dunno. cbrake, rolf and zecke still have access. rpurdie and kergoth are locked. Nov 06 11:29:58 * mickeyl deletes the errornous dev branch Nov 06 11:30:11 strange. will ask the others. thanks anyways Nov 06 11:30:23 may be any ssh brutforcing protection? Nov 06 11:31:18 could be, i don't know whether ubuntu has such facilities Nov 06 11:36:15 mickeyl: hey, i never locked an account (or two) Nov 06 11:36:56 rwhitby: I have one quick question, how is palm reacting to people patching the .js files? will they start to use obfusicating in future releases (it also makes js code faster) Nov 06 11:37:20 rwhitby: ah btw., i sent you a mail a couple of days ago, but never got a reply. Nov 06 11:37:39 zecke: they're fine with it. they even did a demo on stage about how to modify the .js files, and they posted the patch in their developer blog Nov 06 11:38:06 mickeyl: the engineers are not allowed to tell us anything we don't already know. Nov 06 11:38:21 rwhitby: ok Nov 06 11:39:11 zecke: we've had conversations with the developer relations and security groups in Palm, and they all love what webos-internals.org is doing, since we're adding some structure and safety to what would otherwise be the wild west. Nov 06 11:40:23 rwhitby: do you think palm engineers will ever participate in upstream communites? e.g. I have some questions on their webkit patches but I have no idea where to catch the devs Nov 06 11:41:29 zecke: I would say firstly approach the opensource@palm email address (yes, there are real people behind it who actually return emails with useful replies). if that doesn't work, I can try and get a contact. Nov 06 11:42:04 rwhitby: thanks Nov 06 11:42:18 zecke: do you have a Pre? Nov 06 11:42:31 rwhitby: no Nov 06 11:43:03 hmm, my connection died for a while there :( Nov 06 11:43:38 * rwhitby is still waiting for the German DHL to actually give some info for a tracking number. 13 days transit so far. Nov 06 11:43:50 does anyone have good knowledge about the LGPL? Specially § 2.b Nov 06 11:43:53 rwhitby: that sounds bad Nov 06 11:44:19 zecke: stefan posted it, so at least I have a real person to chase it up locally :-) Nov 06 11:44:52 or he kept it and made up DHL number :-D Nov 06 11:45:56 zecke: 2b is routinely ignored for both GPL and LGPL stuff. Nov 06 11:48:08 zecke: what is your issue with 2.b? Nov 06 11:48:32 pb__: I'm upset by how google is handling their webkit for of android Nov 06 11:48:49 what are they doing? Nov 06 11:48:59 pb__: basically they have put svn rxxxx into the git tree, and on releases take the latest source from their perforce tree and dump it into git as one commit Nov 06 11:49:12 Pretty much the same thing as they're doing for everything else, then? Nov 06 11:49:27 pb__: this dump contains a wild mix of backports (without the changelog entries in the original tree) and their changes Nov 06 11:49:35 On the bright side, they're using webkit instead of their own random NIH stuff Nov 06 11:49:57 pb__: I have wasted a day trying to figure out what they changed and I wonder if I have anything to hold against them Nov 06 11:50:17 broonie: well, the should have done NIH... because they have (L)GPL in userspace... they were just too lazy about it Nov 06 11:50:45 zecke: well, unless you are a webkit copyright holder, the answer is definitely that you do not have any cause for complaint. but even if you are an owner of the webkit copyright it doesn't sound as though anything they are doing is contrary to the LGPL. Nov 06 11:50:55 zecke: for the 2b stuff there is an argument that patches provide the prominent modification thing. Nov 06 11:51:24 pb__: oh, their android port is based on files that carry my copyright Nov 06 11:51:53 zecke: The kernel folks appear to be making progress by calling them on their rubbish development practices. Nov 06 11:52:27 pb__: so, it is legal that I start hacking on linux 2.6.31, merge with linux 2.6.32 and provide a diff against 2.4.0? Nov 06 11:52:40 zecke: okay, in that case you obviously are entitled to enforce the license against them. but, based on what you have written above, it doesn't sound like they are actually contravening the licence. Nov 06 11:52:50 zecke: as I understand it, yes, that would be legitimate. Nov 06 11:52:54 pb__: too bad :) Nov 06 11:53:14 the lgpl doesn't contain any obligation to use sensible development practices Nov 06 11:53:31 one of the things you give up by using GPL/LGPL is any real control Nov 06 11:53:35 on hindsight a tremendous lack Nov 06 11:53:56 pb__: I had hoped the 2b would do that... as it forces to put the "date" of the change in it Nov 06 11:53:57 mickeyl: Well, control would result in failure to meet key goals for the license. Nov 06 11:54:10 they could technically publish the changes in BASE 77 dumps and still be obeying GPL/LGPL Nov 06 11:54:11 broonie: no, i meant the the sensible development practices Nov 06 11:54:42 zecke: oh, right, you are talking about LGPL 2.1. I was thinking of 2b in LGPL 3, which is something completely different. Nov 06 11:54:47 zecke: "On this day we copied the entire release X.Y.Z source on top of it". Nov 06 11:55:08 XorA: There is the preferred form of modification thing, at least. Nov 06 11:56:28 pb__: so my understanding is a backport should be marked with a different date than a patch... as the whole webkit Android2.0 didn't happen on one day should have different commits, patches, or marks Nov 06 11:56:35